Improved Quantum Genetic Algorithm And Its Application In Image Segmentation

Image segmentation as a kind of basic computer image processing technology,has always been a research hot topic at home and abroad, it is the basis of image analysis and understanding, and it is also an important link of automated image processing. This paper is based on the improved quantum genetic algorithm, and applies it to threshold segmentation and fuzzy edge detection. The concrete research content is as follows:First, the improvement of traditional quantum genetic algorithm is studied. In order to maintain a good population structure and avoid falling into local optimum,a quantum crossover and mutation operator is introduced in. In order to In order to enhance the global optimization and more quickly locate close to the global optimal solution, a narrow domain policy is used. In addition, a adaptive strategy is used in the rotation angle of the quantum gate to accelerate the convergence. Finally,through the simulation experiment of multi-peak value functions, the performance of the algorithm is verified.Second, the application of threshold segmentation which is based on improved algorithm is studied. In this paper, the improved algorithm is applied to the 2-D Fisher threshold segmentation, through image segmentation experiments, the anti-noise performance between 2-D Fisher algorithm and other 2-D algorithms is compared, and compared with the traditional genetic algorithm and the basic quantum genetic algorithm, the improved algorithm in this paper improves the search efficiency and precision of the threshold vector. In addition, the improved algorithm is also applied to the 3-D Otsu threshold segmentation, the algorithm enhances the computational efficiency, and its good segmentation results is verified through the image segmentation experiments.Thirdly, the application of Fuzzy Edge Detection which is based on improved algorithm is studied. In order to improve noise immunity of general fuzzy edge detection algorithm, this paper presents a two-dimensional histogram fuzzy edge detection algorithm.The method constructs a two-dimensional histogram through pixel neighborhood information and obtains the threshold vector by 2-D Otsu, then it redefines the membership function according threshold vector. In addition, the method combines QGA to quickly search the threshold vector, saving a lot of time.Simulation results show that this method can obtain a better edge detection effect,and also has a stronger anti-noise performance at the same time.